How to train your filters with historical data

Updated 2 months ago by Marie Agard

Riminder is a self-learning platform. Through deep learning algorithms, the system takes into account all your feedbacks on the platform and learns from it. You can train the filters with historical data to get better results that reflect your preferences. After every 10 feedbacks, Riminder platform will trigger a retraining for the job filter. 

Step-1 

Create a folder source on the platform. In that source, add 20 or more CVs of candidates recruited in the past for a particular job.

Step-2 

Once these CVs are scored for that job filter, select the folder source. For each profile, click on” YES” and give a 4-star feedback.

Step-3

Create a folder source on the platform. In that source, add 20 or more CVs of candidates rejected in the past for a particular job filter.

Step-4 

Once these CVs are scored for that particular job filter, click on No for each profile and give a  1-star feedback to each.

You can repeat this process for all the job filters on your platform and the system will retrain itself according to the hiring preferences of your company. When you are using the platform, you can also move profiles to YES, NO or LATER tabs as per your preference, and also give feedbacks to each of the profiles on the platform. This will help the system to improve the scoring in line with your company’s hiring culture. Retraining triggers after every 10 profiles are given feedback for a particular job filter.


How Did We Do?